Contentsquare today published survey data putting a number on what happens after an artificial intelligence assistant sends a shopper to a brand. Only 3% of respondents said they would complete the purchase if the website that followed disappointed them. The remaining 97% said they would keep researching, switch to a competitor, or walk away.

The Paris and New York headquartered digital analytics company released the findings on September 2, 2026, drawn from a survey of 2,000 consumers in the United States and France. The central measurement concerns the handoff: what a consumer does when an AI recommendation arrives ahead of a website experience that fails to support it.

According to Contentsquare, 57% of respondents in that scenario would keep researching, 23% would switch brands, and 16% would abandon the purchase entirely. Those three figures plus the 3% who would buy anyway total 99%, a gap the release does not explain and which is most plausibly a rounding artefact. The company does not itemise it.

What the survey measures, and what it does not

The methodology section is short. Contentsquare surveyed 2,000 consumers across the United States and France through Pollfish, the survey sampling platform, asking about AI use in online shopping, brand discovery and consideration, purchase decisions, agentic commerce and digital experiences. All published figures reflect what the company calls stratified and weighted global results across the combined two-country sample.

That description carries a wording problem worth naming. A sample drawn from the United States and France is not global, and the release does not give the split between the two markets, the fieldwork dates, the margin of error, or the weighting variables. Nor does it separate United States responses from French ones anywhere in the published data, which removes the possibility of checking whether the headline numbers hold across both markets or are driven by one.

A second arithmetic gap appears in the spending data. Contentsquare reports that 57% of consumers say AI recommendations have already moved real money out of their pocket. The breakdown that follows is expressed as a share of all consumers rather than of that group: 21% spent between 50 and 150 dollars on AI-influenced purchases, 14% spent between 150 and 500 dollars, and 5% spent 500 dollars or more. Those three bands sum to 40%, seventeen points short of the 57% headline. The most likely explanation is a band below 50 dollars that the release omits, but the company does not say so.

None of this makes the direction of the finding implausible. It does mean the numbers arrive without the scaffolding an analyst would need to interrogate them.

Trust in the recommendation, doubt about the source

The trust measurements are the part of the release most likely to be quoted out of context. Two-thirds of respondents, 67%, said they would trust an AI recommendation over that of a social media creator or influencer when evaluating an unfamiliar brand. Nearly half, 47%, said AI has changed their mind about which brand to buy.

Both are relative measurements. They rank AI against a specific alternative rather than establishing how much confidence consumers place in AI outputs on their own terms. Research published this year has repeatedly found the absolute figure to be low. Yelp and Morning Consult, surveying 2,202 United States adults, found that only 15% trust AI search platforms a lot even though 65% had used one in the previous six months, with 63% double-checking results elsewhere. Reddit's Path to Purchase research found half of United States shoppers verifying AI product recommendations on the platformbefore completing a purchase.

RTB House reached a similar structural conclusion in August, reporting that leading AI assistants now outrank TikTok, Instagram, Facebook, newspapers and influencers on shopping trust while 42% of United States respondents said the same tools lengthen the time they need to settle on a purchase. Reddit's own comparison found influencer reviews trailing peer posts by roughly two to one in United States buying decisions, which suggests the influencer benchmark Contentsquare used is a low bar in the first place.

Read alongside those studies, the Contentsquare figure describes a hierarchy of scepticism rather than an endorsement.

Where AI now sits in the shopping journey

Thirty-six per cent of respondents said AI has replaced traditional search engines for at least some shopping activity. Within that behaviour, 53% use AI to find the best price or deal, 52% use it to compare products, and 29% use it to discover new brands.

Measured adoption figures across the industry land in a wide band, and the spread depends heavily on question wording. Smarty Marketing's survey of 1,295 United States consumers found Google holding 56.68% of shopping search starts with ChatGPT at 7.26%, a much narrower picture of platform preference than self-reported usage produces. Optimove's holiday research put 72% of United States consumers consulting AI assistants for gift ideas. NIQ, publishing days before the Contentsquare release, catalogued the same dispersion across comparable studies. Asking whether AI has replaced search for some activity is a lower threshold than asking where a shopping session begins, and it returns a correspondingly higher number.

The generational pattern in the Contentsquare data cuts against the usual assumption. Millennials, defined here as consumers aged 30 to 44, lead adoption at 74% likely to use AI when shopping online, ahead of Gen Z at 70% for ages 18 to 29 and Gen X at 65% for ages 45 to 60. Millennials also lead on brand discovery at 35% and on price and deal hunting at 57%. The release does not report figures for consumers over 60.

The consistency problem

The finding with the clearest operational consequence is smaller than the headline. About one in five respondents, 21%, said an AI shopping assistant gave them information that differed from what they subsequently found on the brand's website. Twenty per cent said AI did not provide enough detail to buy with confidence.

That is a data accuracy question wearing a customer experience costume. When an assistant states a price, a specification, a shipping window or a returns policy that the destination page contradicts, the discrepancy surfaces at the moment of highest purchase intent.

Independent measurement supports the concern. An audit published on 18 August 2026 by the research firm Empirank examined 1,257 testable factual claims across 182 businesses recommended by a search-enabled OpenAI configuration and found that twenty of those businesses, or 11.0 percent, carried at least one claim the evidence contradicted, with a further 479 claims the researchers could not resolve either way. A separate study covering 4,776 audited food venues in Bali found ChatGPT and Gemini missing 85.6% of them entirely. Earlier research found 47.1% of marketers encountering AI inaccuracies several times each week.

Brands do not control what an assistant says about them. They do control what the destination page says, and the Contentsquare data measures the cost of a mismatch between the two.

The quotes

Jean-Christophe Pitié, chief marketing officer at Contentsquare, framed the finding as a question about where optimisation budgets end up.

"As companies pour money into AEO and GEO to make their brands visible to AI, the next question is: what happens after the AI sends the customer your way?" said Pitié. "Visibility only creates value if it converts or drives loyalty, translates into action, and that ROI ultimately lives in conversion. If the experience that follows is slow, confusing or frustrating, no amount of AI-driven discoverability will deliver the return businesses expect. The companies closing that gap are the ones building customer confidence through great end-to-end experiences."

He returned to the consistency point in a second statement.

"Brands have spent years trying to create a single source of truth for customers; AI just created another one," Pitié added. "Consumers are forming opinions based on what AI tells them long before they reach a website. Brands need to make sure that information is consistent - and that the website experience proves the recommendation right, not wrong. That's the real proving ground of AI business value."

The commercial interest behind that framing is not concealed. Contentsquare sells experience analytics: session replay, heatmaps, zone-based analysis and frustration scoring across web, mobile and app channels, on a platform the company says more than 1.3 million websites use. Research concluding that post-click experience determines conversion outcomes maps directly onto what the company sells. That does not invalidate the data. It does mean the data was commissioned by a party with a position in the answer.

Why this matters for marketing teams

The budget question Pitié raises has been building through 2026 without a settled answer. Adobe research covering more than 500 marketers found 98% lacking a confident AI search roadmap. Fractl's survey of 343 United States marketing decision-makers found 81% still saying SEO when discussing AI search strategy internally, with only 19% having adopted the GEO label at all. Vendor tooling has moved faster than the vocabulary: HubSpot shipped a dedicated answer engine optimisation product in April, disclosing at the same time that organic traffic for its own customers had fallen 27% year over year.

The evidence underpinning that spending remains contested. A critical survey of 45 studies posted to arXiv in July concluded that no reviewed generative engine optimisation technique produces a stable cross-platform effect on discoverability or downstream traffic, with one rewrite scenario cutting a page's retrieval by 16%.

Against that backdrop, the Contentsquare release argues for a measurement boundary rather than a channel. Visibility inside AI Overviews, AI Mode or a chatbot answer is an input. What the survey measures is the conversion event that sits several steps downstream, on infrastructure the brand still owns.

There is a retention dimension too. Adobe research published in August found 72% of shoppers deleting retail apps after a single use, and Optimove's holiday data found 53% of United States consumers intending to buy only from retailers they used the previous year. A first impression formed inside an assistant and contradicted on arrival is a poor foundation for either.

Contentsquare's positioning

The research fits a sequence of moves the company has made through the past nine months. Contentsquare extended session replay, heatmaps and frustration scoring into Shopify checkout flows in a partnership announced on December 17, 2025. It connected its behavioural datasets to Dust's enterprise agent platform through a Model Context Protocol connector on June 24, 2026. It pushed behavioural segments directly into Klaviyo through a native integration announced on July 16, 2026.

The category carries regulatory exposure in one of the two markets surveyed. France's data protection authority CNIL opened a public consultation on February 25, 2026, on a draft recommendation governing session replay tools, which would require prior consent as a distinct purpose for the technology that underpins much of the analysis Contentsquare sells.

Contentsquare made the research available to PPC Land under embargo on August 31, 2026, with publication set for September 2. The company has not published the underlying questionnaire, the response-level data, or a breakdown by market.

Timeline

Summary

Who: Contentsquare, the Paris and New York based digital experience analytics company, published the research. Jean-Christophe Pitié, its chief marketing officer, provided the accompanying statements. Fieldwork ran through Pollfish. The findings concern retailers, e-commerce teams, conversion specialists and the agencies selling answer engine and generative engine optimisation services.

What: A survey reporting that only 3% of consumers would complete a purchase when an AI recommendation is followed by a disappointing website experience, while 57% would keep researching, 23% would switch brands and 16% would abandon the purchase. Related findings include 67% trusting an AI recommendation over a social media creator when evaluating an unfamiliar brand, 47% saying AI changed their mind about which brand to buy, 36% saying AI has replaced traditional search for at least some shopping activity, 21% reporting information from an AI assistant that differed from the brand's website, and 20% saying AI did not supply enough detail to buy with confidence.

When: Contentsquare published the research on September 2, 2026, having offered PPC Land an early look under embargo on August 31. Fieldwork dates are not disclosed.

Where: The sample covers 2,000 consumers in the United States and France. Results are reported only in combined weighted form, with no market-level breakdown.

Why: Optimisation budgets have shifted toward making brands visible inside AI answers, while the evidence base for those techniques remains contested and the accuracy of AI recommendations remains uneven. The survey quantifies the step after visibility, measuring what consumers do when the assistant's account of a brand and the brand's own site fail to agree. The findings arrive from a vendor that sells tooling for the problem they describe, and the release omits the fieldwork dates, market split and margin of error that would allow independent scrutiny.